CSIR-NCL Celebrates National Technology Day 2025
Pune, 16 May 2025: CSIR-National Chemical Laboratory (CSIR-NCL), Pune celebrated National Technology Day with great enthusiasm. The theme for this year, YANTRA—Yugantar for Advancing New Technology, Research, and Acceleration—highlighted the nation’s commitment to self-reliant technological innovation on global scale.
Dr. Ashish Lele, Director, CSIR-NCL, extended a warm welcome to the chief guest Mr. Karthick R., Chief Manager (Innovation centre), ONGC-MRPL, Mangalore and addressed the audience with a brief background on the significance of National Technology Day and also highlighted the importance of integrating emerging technologies like AI and ML in the industrial sector.
Chief Guest Mr. Karthick delivered his talk on the topic entitled “Deployment of Artificial Intelligence and Machine Learning Applications to Improve Productivity in Petroleum Refining and Petrochemicals Industry.”
In his talk, Mr. Karthick R. presented a detailed account of how AI and ML are being leveraged at MRPL to drive operational efficiency. He opened by addressing the challenge of managing vast amounts of industrial data, quoting John Naisbitt: “We are drowning in information but starved for knowledge.” He shared insights from MRPL, stating, “If knowledge is power, data is knowledge.” He discussed how big data computing at scale is enabling the development of dynamic, data-driven industrial applications. He elaborated on different types of algorithms and explained the AI/ML Algorithm Tree, offering a foundational understanding of how these tools are structured and used in real-time scenarios.
A major focus of his presentation was MRPL’s real-world application of AI and ML in addressing an industrial issue. MRPL had been experiencing odour problems in the final propylene product. To tackle this challenge, the MRPL Innovation Center turned to artificial intelligence and machine learning techniques. By adopting a data-driven approach, they were able to identify and resolve the underlying causes of the issue. The combination of Explainable AI and association mining revealed critical operational parameters. The novelty of this initiative lies in its patented models and hybrid algorithms. Mr. Karthick compared mechanistic models, such as kinetic first-principle models, with AI/ML-based models, offering a clear distinction in their design and functionality. He also discussed Real-Time Optimization systems and informed about the Real-Time Condition Monitoring App developed at MRPL. He described its features and model performance, emphasizing its role in optimizing industrial productivity and decision-making.
The lecture concluded with an engaging Question & Answer session and a vote of thanks, bringing the event to a successful close.
